Video motion analysis with automated tracking: an insight

Abstract

The article describes the use of elementary techniques in computer vision and motion photography for the analysis of well known experiments in interactive instructional physics laboratories. We describe a method for the automated tracking of the kinematics of physical objects which involves the subtraction of orthogonal colors in color space. The aim is to expose undergraduate students to image processing and its applications in video motion analysis. The straightforward technique is simple, results in computational speedup compared to an existing method, removes the need for a laborious repetitive and manual tagging of frames and is generally robust against color variations. Insight is also presented into the process of thresholding and selecting the correct region out of the several choice presented in the post-threshold frames. Finally, the approach is illustrated through a selection of well known mechanics experiments.

‘PhysTrack’: a Matlab based environment for video tracking of kinematics in the physics laboratory

“…It then detects these sections in each individual frame and determines the trajectory of each connected region throughout the video. This algorithm has been described previously [16]. Table 1 compares main features of the two types of tracker.…”

“…Matlab [15] is a popular numerical and graphical integrated development environment equipped with numerous toolboxes making it a favourite choice for educationists around the globe. Unsurprisingly, Matlab has been extensively used in video motion tracking as well [6,16,17]. The software provides an impressive array of analysis tools and a programming language similar to C. However, performing video motion tracking becomes tiresome due to the complexity of the language skills required.…”